Let's talk about AI and Bridgeton, New Jersey….

39 pensamientos sobre “Let's talk about AI and Bridgeton, New Jersey….”

1. AI freaks me out.

2. Winston Smith's job is at risk.

3. AI is garbage and dangerous.

4. I know Bridgeton, I find it easily to believe there are murders in Bridgeton almost any day of the year. If that event didn't occur in Bridgeton I guarantee there were plenty of events just like it that did !

5. It belongs in news rooms just like calculators belong in engineers pockets… amazing tool, but garbage in is garbage out and only those that know how to do it without AI are the ones that can actually use AI to its full potential, using AI to augment your job can be really really time saving, but using it wrong or trying to get it to do a qualified persons job… won't end very well, as can be seen here.

6. "Artificial Imagination"?

7. Not gonna lie, I thought the word in the title of the video was Bridgerton, not Bridgeton.

8. God, this disclaimer makes me so mad. It is the journalist's job to "verify the critical parts with trusted sources." That is most of their job.

9. AMERICA CAN NOT DEAL WITH A CON PRESIDENT(trump),that’s why the law is so slow.

10. We have fully transitioned from the information age to the misinformation age.

11. The moment a news outlet uses AI in it's reporting I'll find a different news outlet.
    I don't have a problem with AI, I have a problem trusting the people using AI to write for them.

12. Thank you for making the point that generative AI is not always right

    People think that generative AI is like a search engine, which could not be farther from the truth

    Generative AI generates stuff, which may or may not as you point out have anything to do with reality

    More people need to get this message so they quit thinking that generative AI is some kind of search engine and actually understand that it’s just making stuff up

13. It's not "artificial intelligence", it's an "automated BS machine". It's a really finely honed producer of BS, and once you understand that about it, you will understand how much to believe a word it says.

14. How many people you think are google the game telephone????

15. I am convinced that the greatest educational failure of the modern age is critical thinking skills. The way the internet has changed how information is shared has caused critical thinking skills to be easily a whole magnitude more important than they were 25 years ago and nobody prepared for it. I don't think fascism would be getting this extra shot that emerged for it had we been more ready for the mass distribution of misinformation and emergence of the echo chamber 'phenomenon'. Perhaps I should say social failure, because while the root is an education shortcoming the effect is felt across society in significant ways.

16. "If you were to try to get it to write an article saying you want something unique about a crime or something like that, it may just literally make one up." AI = Fox News. Got it.

17. There’s already enough artificial intelligence out there, now 50-60% of people have no clue what is going on and will believe almost everything.

18. The news is supposed to be the trusted source!

19. Have we learned nothing from Star Trek? AI is a co-pilot, not a driver. It's there to assist not so much to create.

20. What worries me is how AI is becoming synonymous with another initial: BS.

21. So basically treat it like Fox News.

22. I listen in a meeting that powerful AI can replace many low and mid tier positions thus saving the organization money. When ask to raise any concern/question, i ask does this include you first in the organization when we transition to AI. Hmm…wonder how all those replace will feed themselves and do actual contributions to the economy.

23. THAT IS WHAT FOX IS DOING!!!

24. Hello everyone.?

25. Hello Mr.Beau.?

26. The best way to understand the AI that writes stuff is to look at the words your phone recommends as you type. THAT is what it is doing. Clicking the very first word recommended, every single time. It will make a sentence. It will be grammatically correct. But, "Crayons have descended to rule jelly bean mountain." is a correct sentence, it means NOTHING.
    They are very fancy "guess the next word" engines. The control you have is to point it in a direction. The most advanced AI available today is still like a toddler with Alzheimer's and perfect diction operating a howitzer. It cannot REMEMBER anything well enough to even be capable of "forgetting" what it was talking about.

27. Frankly, it frightens me that people think Large Language Models can be interchangeable with AGI (and more importantly, interchangeable with those pesky, paycheck-wanting human workers! ) just because they can pass Turing tests. A Large Language Model (what most people think of when "AI" is mentioned today) has no concept of truth and falsehood, or even that objects exist or don't exist.
    Their function is to produce text that looks like the most plausible response (or rather, statistically one of the most plausible responses) to an input. Writing a news article that looks like any other news article, about a crime like any other IS success to such a model. It has no idea that some crimes even happened and others didn't!
    An LLM has no concept of 'reality', only 'plausibility'. They're fundamentally misapplied attempting to produce factual output, because they have no concept of facts.

28. The greed of corporations to eliminate jobs is very damaging to all of us.

29. We also called the game of Telephone, Gossip. I'm in Mid-Michigan.

30. It's not even really AI, it's a turbo version of autocorrect! It's literally just predictive text with a massive data set to draw on!

31. AI doesn’t differentiate between fiction and nonfiction fiction where it puts together a narrative, AI is perfect for producing propaganda and media companies will use it to increase profits less use of humans.

32. For what feels like ages I have told to whoever I was talking with that AI is more of a danger than it is a benefit. Several texting apps on my phone have added some version of AI to their products but, while I can't remove that component, I have no intention of using it. It's basic self defence.

33. the issue is that these ai arent what people think they are. they arent talking and they arent trying to convey any ideas, they dont learn to tell the truth, they learn about words and sentances and language. in other words all theyre doing is producing writing that LOOKS like other writing it was shown, its grammatically correct, but any information conveyed, whether true or false, is irrelevant because its goal is to be able to sound like a human being, not tell the truth

34. So…the story was written by Fox News?

35. Another example of AI hallucinations again.

36. The 60 Minutes episode on AI is great. It begins with, "AI is not as good as you'd hoped, but not as bad as you'd feared." One day maybe. But today this is the most accurate description of AI.

37. Reminds me of an article I read about one AI text program that happily wrote articles on flux capacitors, the history of bears in space flight, and how the Harry Potter book series is about the gay romance between Harry and Ron. People talk about AI as being some revolutionary technology, when it isn't even smart enough to determine when something is obviously false.

38. Honestly, I skimmed the description and thought, Beau's a fan of Bridgerton? Wow! ??
    I need to get my eyes checked.

39. There is a term for this. It's called an AI hallucination and OpenAI is being sued for defamation for one of the more egregious instances where a journalist trying to research a trial was given a result that said he had embezzled money and was being tried for fraud. To paraphrase the movie Broken Arrow, I don't know which is more frightening, the fact that this happened or that it happens so often that the industry has given it a name.

Los comentarios están cerrados.